Based on pilot work in the states of TX and PA, the purpose of this Collaborative R34 for Pilot Studies of Innovative Treatments application is to prepare for a multi-site, Phase III clinical trial of a Self-Directed Care (SDC) intervention for people with serious mental illness (SMI). In SDC, participants manage a personal budget from which they purchase goods and services including specific types of health care, social supports, and items that allow them to live independently. Because the amount of money that participants are allocated is equal to the amount that would have been spent on serving them in traditional programs, the approach is designed to be budget neutral. Evidence from a three-state randomized controlled trial with elderly and disabled Medicaid beneficiaries found that, compared to services as usual, SDC participants had better health outcomes, higher satisfaction with services, and fewer unmet needs. Yet despite this compelling evidence, use of this approach for people with SMI has been limited. With their consent, 88 subjects at each of the two sites will be randomly assigned to SDC or to services as usual and followed for twelve months. SDC participants will be helped to develop a person-centered recovery plan and corresponding budget. Upon approval of their budget, participants will use budgeted funds to purchase mental health and other services and supports. The primary outcomes will be symptom reduction, enhanced functioning, and reduction of unmet needs, with secondary outcomes of recovery and behavioral activation for self-management of health and mental illness. The study has four aims. The first aim is to develop and refine an SDC intervention implementation manual as well as a fidelity assessment instrument and procedures to measure staff's adherence to the model. The second aim is to conduct a small randomized pilot test of the effectiveness of the SDC intervention in order to evaluate all research procedures. The third aim is to assess SDC intervention acceptability, test for differences in service satisfaction by study condition, and determine reasons for study attrition among those assigned to both conditions. The fourth aim is to examine the validity, reliability, and suitability of local mental health authority administrative data for use in utilization and cost analyses for a future R01 study. To guide this multi-site study, an eight-member project Steering Committee will be formed, comprised of each site's Principal Investigator, Project Director, local mental health authority representative, and consumer consultant. The committee will be charged with the scientific integration of research procedures, training of research and intervention staff, fidelit monitoring, and development of a publication policy. Both sites will have access to the project's de-identified cross-site data base and will use the same analytic tools in a coordinated application of hierarchical, mixed-effects multivariable statistical models.